ADT Use and Nodal Irradiation in Men Receiving Post-Prostatectomy Salvage Radiotherapy within a Statewide Radiation Oncology Quality Consortium

PURPOSE/OBJECTIVE(S)

For men with biochemical recurrence after radical prostatectomy, salvage radiotherapy (SRT) is a standard of care. Outcomes are improved when SRT is delivered at lower PSA levels, and there has been increased emphasis on more timely treatment. With early SRT, however, there remains uncertainty as to the optimal use and duration of androgen deprivation therapy (ADT) and pelvic lymph node radiation (PLNRT). Moreover, PET imaging and genomic classifiers have emerged as tools to guide treatment decisions, but their uptake in routine practice is unknown. To address these questions, we analyzed a contemporary cohort treated with SRT within the Michigan Radiation Oncology Quality Consortium (MROQC). We hypothesized that ADT and PLNRT practices would reflect recent trial results in this setting.

MATERIALS/METHODS

Eligible patients receiving SRT at an MROQC center were enrolled from 06/09/20 to 11/04/22. Data was prospectively collected via patient-, physician-, and physicist-completed forms. Patients were matched to the Michigan Urological Surgery Improvement Collaborative (MUSIC) database for additional treatment- and patient-related data. Univariable (UVA) and multivariable analyses (MVA) were performed to test associations between patient/tumor factors and ADT or PLNRT use.

RESULTS

A total of 191 patients across 26 centers were enrolled in the MROQC database. Of these, 116 were matched to the MUSIC database. Median time from RP to SRT was 17 months (IQR 8 - 33 months). The median post-RP PSA prior to SRT was 0.25 (IQR 0.16 - 0.60). Early SRT was defined as pre-SRT PSA ≤0.5, and 27% (n = 31/116) had a pre-SRT PSA >0.5. Twenty-eight were pT3b/T4, 97% were pN0/NX, and 51% had positive surgical margins. Fractionation was conventional (>28 fractions) in 58% and moderate hypofractionation (20-28 fractions) in 38%. Table 1 describes the patients receiving ADT and/or PLNRT. Median ADT duration was 6 mo (IQR 6 - 7 mo). MVA revealed pre-SRT PSA >0.5 (OR 5.05 [1.89 - 15.33]) and pT3b/T4 disease (OR 4.23 [1.40 - 14.56]) were significantly associated with ADT use (p <0.05), but not grade group (GG) or margin status. PLNRT was significantly associated with pre-SRT PSA >0.5 (OR 3.04 [1.21 - 8.42], p <0.05) but not pT stage, margin status, or GG. PET imaging was performed in 37% of men (52% negative, 21% prostate bed alone uptake, and 26% lymph node positivity) and genomic classifiers were performed in 24%.

CONCLUSION

Nearly 75% of biochemically recurrent prostate cancer patients within MROQC received early SRT, and about half received ADT. A pre-SRT PSA >0.5 was strongly associated with ADT and PLNRT. With prostate bed SRT alone, very few received ADT. Given the considerable heterogeneity in treatment, additional studies may help identify patients who most benefit from ADT + PLNRT, and who may be spared potential added toxicity.

Cannabis Use Patterns among Patients with Early-Stage Breast Cancer in a Large Multicenter Cohort from a State with Legalized Adult Non-Medical Cannabis

PURPOSE/OBJECTIVE(S)

Cannabis use among patients with cancer is an area of great interest given its widespread acceptance despite the lack of supporting clinical data. The absence of data limits the understanding of potential clinical benefits of cannabis and the ability of providers to deliver evidence-based recommendations for patient care. We explored cannabis use patterns in patients with early-stage breast cancer in a large multicenter cohort in a state with legalized adult non-medical cannabis.

MATERIALS/METHODS

Initial questions about cannabis use history and frequency were introduced in Michigan Radiation Oncology Quality Consortium (MROQC) breast cancer patient surveys on 2/1/2020 for female patients receiving radiation after lumpectomy for non-metastatic breast cancer. Expanded questions were introduced on 6/28/2022 to assess mode of administration, active ingredient, and reason for use. Summary statistics were generated. A multivariable model using logistic regression identified patient characteristics associated with cannabis use.

RESULTS

Among 3948 eligible patients, 2738 (69.35%) completed survey questions, and 2462/2738 (89.9%) completed the initial question on cannabis use. Among those, 364/2462 (14.8%) noted cannabis use in the last 30 days, 588 (23.9%) noted remote use (>30 days ago), 1462 (59.4%) reported never having used cannabis, 44 (1.8%) preferred not to answer cannabis use questions, and 4 (0.4%) did not provide use history. Younger age [age <50 vs 60-70, OR 2.5 (95% CI 1.65, 3.79) p<0.001)], Hispanic ethnicity [OR 2.20 (95% CI 1.06, 4.56) p = 0.03], history of smoking [OR 2.56 (95% CI 1.88, 3.48) p<0.001], current smoking [OR 4.70 (95% CI 3.22, 6.86) p<0.001)], and prior chemotherapy [OR 1.40 (95% CI 1.00, 1.96) p = 0.05] predicted recent cannabis use in a multivariable model. Of the 364 patients endorsing cannabis use in the last 30 days, 89 (24.5%), 72 (19.8%), 29 (8.0%), 66 (18.1%), 30 (8.2%), and 78 (21.4%) reported using cannabis 1-2 days, 3-5 days, 6-9 days, 10-19 days, 20-29 days, and all 30 days, respectively. The most common modes of administration among 76 individuals who responded to the expanded questionnaire to date were oral (39.4%), smoking (30.3%), and topical (10.5%). The products used contained tetrahydrocannabinol (THC; 26.3%), cannabidiol (CBD; 19.7%), balanced levels of THC and CBD (19.7%), or active ingredients that were unknown to the patient (34.2%). Patients frequently endorsed cannabis use for insomnia, anxiety, and pain.

CONCLUSION

Many patients with early-stage breast cancer are using cannabis. Younger age, Hispanic ethnicity, smoking, and chemotherapy history are predictors of cannabis use. Patients are often unaware of the active ingredients in the products that they use, suggesting an important role for patient education and a need to equip providers to advise patients in their care.

Androgen Deprivation Therapy Use among Intermediate Risk Prostate Cancer Patients Undergoing Radiation Therapy across a Statewide Radiation Oncology Quality Consortium

PURPOSE/OBJECTIVE(S)

For men with intermediate (INT) risk prostate cancer, the addition of androgen deprivation therapy (ADT) reduces risk of PSA failure, distant metastasis, and cancer-related mortality. Moreover, the relative reduction in cancer-related adverse outcomes with ADT use appears consistent across all INT risk subgroups. The absolute benefit of ADT, however, varies by baseline risk. In contemporary practice, it is unknown which clinical factors are most strongly associated with intended ADT use. Therefore, we sought to identify such factors within the diverse practices of the Michigan Radiation Oncology Quality Consortium (MROQC).

MATERIALS/METHODS

Patients with localized prostate cancer undergoing definitive radiation therapy were enrolled from 6/9/20 to 11/4/22 (n = 599). Standardized patient, physician, and physicist forms were used to collect baseline and follow-up information. Intended ADT use, defined by the treating physician, was prospectively collected and is the primary outcome of this analysis. Univariable (UVA) and multivariable analyses (MVA) associations with patient (age, race, comorbidities), tumor (T stage, Gleason, percent cores positive, and PSA), and practice-related (academic vs private) factors were performed. In addition, advanced modality testing (PET, MRI, and genomic classifiers) was available as of March 2021, and subgroup analysis were performed where appropriate.

RESULTS

A total 351 patients across 26 centers were enrolled with INT risk disease. ADT use was intended for 46% of men (n = 162/351) which differed by men with NCCN favorable INT (21%, n = 22/105) vs unfavorable INT risk disease (57%, n = 140/246), p<0.001. Sixty two percent (n = 100/162) had an intended ADT duration of 4-6 months and 21% (n = 34/162) had ≥12 months. Older age was associated with ADT use (70 vs 67, p < 0.01); there were no significant differences by race or comorbidities number. MVA showed Gleason 4+3 (OR 4.61 [2.91 - 7.42]) and > = 50% positive cores (2.56 [1.52 - 4.37]) were significantly associated with ADT use. No significant differences were noted based on practice setting. Pelvic MRI was obtained for 71% of men (n = 197/279), genomic classifiers in 47% (n = 130/279), and PET in 2% (n = 6/282). In the subset with MRI (n = 197), adverse features (ECE, SVI, or equivocal LNs) were associated with intended ADT use (OR 3.0 [1.4 - 7.1]) after adjustment for NCCN favorable/unfavorable INT risk classification.

CONCLUSION

Within a state-wide consortium, intended ADT use for intermediate prostate cancer is most strongly associated with Gleason score, ≥50% positive cores, NCCN unfavorable intermediate risk classification, and adverse features on MRI. Nearly half of men had genomic classifier testing underscoring the importance ongoing trials such as NRG/GU 010.

Use of Explainable AI Algorithm Revealing Longitudinal Changes in Practice Patterns and Toxicity Models

PURPOSE/OBJECTIVE(S)

Dosimetric constraints evolve as clinicians implement practice changes, requiring modeling approaches to be dynamic. We applied a semi-automated explainable artificial intelligence (eAI) algorithm and dashboard visualizations to model dysphagia and xerostomia for head and neck cancer patients. We coupled a large, comprehensive, "real-world" database to the eAI for discovery of features with the strongest combined statistical and machine learning based evidence and to identify clinically actionable thresholds.

MATERIALS/METHODS

Cohort included 758 patients treated 2017-2021 for HN cancer with conventional fractionation. Features included age, sex, diagnosis, staging, chemotherapy, smoking and alcohol status, BMI, weight loss, re-simulation, DVH curves, PTV and OAR volumes. Patients were scored for toxicity within 2 yrs of RT for dysphagia grade ≥ 3 and xerostomia grade ≥ 2. Bootstrap resampling of thresholds, ROC-AUC, PR-ROC, SN, SP, F1 and diagnostic odds ratio was used to statistically profile strength of evidence for candidate features. XGBoost models with 10-fold cross validation were repeated (n = 20) to identify mean and CIs for statistical measures of predictions. DVH metrics included standard template values and those with highest statistical evidence and low cross correlation with other features. Backward feature selection was used to identify the most relevant feature subset, where the least informative feature is iteratively removed from the model. This workflow was repeated by year and overall.

RESULTS

Annual incidence of dysphagia averaged 0.13 ± 0.02 overall years. Xerostomia incidence decreased from 0.32 to 0.12 (2017-2021). Box-whisker plots by year showed consistent reductions in standard practice toxicity linked DVH metric values. Median dose to superior constrictors (PCM), contralateral parotid and contralateral submandibular gland (SMG) declined from 2017 to 2021 by 48 to 33 Gy, 17 to 10 Gy, and 28 to 22 Gy respectively. Statistics of XGBoost models of dysphagia for all years were ROC-AUC = 0.72 ± 0.05. Strongest overall years predictors were Oral Cavity (OC) D50%[Gy] < 32, inferior PCM Max [Gy] < 60, contralateral SMG D10%[Gy] < 53 and use of Paclitaxel. Xerostomia models were less predictive with ROC-AUC = 0.65 ± 0.05. Strongest predictors over each year were ipsilateral parotid D30%[Gy] < 35, contralateral SMG D96%[Gy] < 18.4, and overall staging < II. Predictive features varied substantially by year for both, showing the most consistency for SMG doses. For example, OC D50%[Gy] < 27 and contralateral SMG D96%[Gy] < 18 dominated xerostomia model in 2017 but not in 2021 when practice norms shifted to lower doses.

CONCLUSION

As OAR doses were systematically reduced, statistical and AI models evidence highlighted contralateral SMG dose as important to both dysphagia and xerostomia for clinical practice change. The "real-world" database + eAI + visualization dashboards provided a method for continuous learning as clinical practice changes.